Journal
ofSterotd
Biochemisrry,
1977, Vol. 8. pp. 847-851. Pergamon Press. Printedin Great Britain
EVALUATION OF RADIOIMMUNOASSAY ALDOSTERONE IN URINE AND PLASMA OF RATS*
FOR
MATHIAS HILFENHAUS Institut
fur Pharmakologie
der Medizinischen Hochschule Hannover, Karl-Wiechert-Allee D 3000 Hannover 61, Germany
9,
SUMMARY
A method is described for the measurement of aldosterone in urine and plasma of rats and the sensitivity, specificity, accuracy and precision of the assay are assessed. The method includes one extraction step by methylene chloride, purification of the extract by one thin layer chromatography and determination of the purified aldosterone by radioimmunoassay. [3H]-Aldosterone is used for the radioimmunoassay as well as for the determination of the recovery of the individual sample. A modified calculation of the radioimmunoassay results is described to correct for the varying amounts of [3H]-aldosterone of the recovery preexisting in the unknowns but not in the standards of the calibration curve. Normal values of plasma aldosterone concentration in adult male Sprague-Dawley rats (23-288 pg/ml depending on the time of day) as measured with our assay are in good agreement with the values reported in the literature. Normal values of urinary aldosterone excretion rate in these rats (3.95.8 ng/24 h) are much lower than those reported in the literature possibly due to the higher specificity of the present assay.
INTRODUCTION
In the rat, in uiuo studies on aldosterone
have been restricted for a long time to the determination of aldosterone secretion in surgically stressed animals (for review see Cl]). Due to the development of highly sensitive radioimmunoassays (RIA) the determination of physiological concentrations of aldosterone in the blood of intact, i.e., conscious and unstressed rats has become possible [l-6]. We recently also reported on aldosterone excretion rates in the urine of rats [7,8]. Since data on urinary aldosterone in the rat as measured by RlA have not been published yet, we herein describe a RIA for the measurement of aldosterone in urine and plasma of rats using extraction and thin layer chromatography (t.1.c.) for separation of aldosterone. This assay has been used routinely in our laboratory during the last 4 years and has proven to be comparatively rapid and simple. MATERIALS
AND METHODS
Animals. Male Sprague-Dawley rats (20&250 g) were kept in a room with controlled temperature (23-24”) and humidity (>SO%) under a lighting regime of 12 h light (06.0&18.00) and 12 h dark (18.0&06.00). The animals were maintained on a standard diet (Na+ 96 mmol/kg, K+ 143 mmol/kg; Altromin. Lage, Germany) and tap water ad libitum or on * Supported by the Deutsche Forschungsgemeinschaft (Sonderforschunasbereich 146 “Versuchstierforschuna”. G 5). Requests f& reprints should be addressed to: Dr: M. Hilfenhaus, Institut fur Pharmakologie, Medizinische Hochschule Hannover, Karl-Wiechert-Allee 9, D 3000 Hannover 61, Germany.
a low-sodium diet (Na+ 4.4mmol/kg, K+ 60mmol/kg; Altromin) and demineralized water ad libitum. For determination of plasma aldosterone rats were decapitated without anaesthesia and blood was collected from the trunk using EDTA (1 mg/ml) as. anticoagulant. Plasma was separated by centrifugation at 0” and stored at -20” until assayed. For measurement of urinary aldosterone excretion rats were placed in individual metabolic cages made of Plexiglas with free access to food and water, the urine--separated from feces and spilled food-was collected in 25 ml-Erlenmeyer flasks. After that urine samples were stored at -20” until assayed. Plasma corticosterone concentrations were identical in rats kept in usual cages or kept in metabolic cages indicating that sitting in a metabolic cage was no stress for the animals. Materials. Methylene chloride, benzene, acetone (all technical grade) were obtained from Merck (Darmstadt) and distilled over a 120cm packed column filled with wire-mesh rings. Toluene (technical grade, Merck) and Triton-X-100 (technical grade, Serva, Heidelberg) were used without further purification. Precoated t.1.c. plates containing a fluorescence indicator (DC-Alufolien, Kieselgel F 254, Merck) were used as purchased. Chromatography was performed in the solvent system benzene-acetone-H,0 (3:2:0.025 by vol) (system A) or in the system ethylacetate-acetone-H,0 (6:1:0.025 by vol) (system B). [1.2.6.7-3H]-Aldosterone (S.A. 103 Ci/mmol, Radiochemical Centre, Amersham) was stored in benzene:ethanol (9: 1 by vol) at 4”. Radiochemical purity was checked by t.1.c. at 3 months intervals. A measureable decomposition of 5% was observed after 6 months of storage and the tracer was purified by t.1.c. 84-l
848
MATHW HILFENHAUS
using solvent system A. o-aldosterone was kindly supplied by Ciba and was used without further purification. The antiserum against aldosterone-21-hemisuccinate was raised in a sheep [lo]. Borate buffer (0.05 M, pH 8), y-globulin buffer (borate buffer containing 1% of bovine immunoglobulin, Behring, Marburg) and Dextran-coated charcoal (6g Norit A, Serva, and 0.2 g Dextran T 70, Pharmacia, Uppsala, suspended in lOOm1 of distilled water) were stored at 4” and were stable for at least 1 month. The scintillation fluid consisted of 40g PPO (2.5-diphenyloxazole, Merck), 1 g POPOP ‘(l&bis-2-(5phenyloxazolyl)-benzene, Merck), 2 1 of Triton-X-100 and 8 1 of toluene. Extraction and purijkation of plasma and urine samples. 0.05-4 ml of plasma or urine were mixed with
known amounts of [3H]-aldosterone (about 5000d.p.m. = 8 pg) as internal standard for determination of individual recovery. After an equilibration period of 12 h the samples were extracted with 20ml of methylene chloride by magnetic stirring for 15 min. The aqueous layer was discarded and the filtered organic extract evaporated. The dry extracts were dissolved in acetone (3 x 0.2 ml) and transferred to the t.1.c. plates using a Desaga-autoliner. The plates were developed in the solvent system A. Aldosterone was located by running spots of unlabeled aldosterone (about 0.5 pg) with each plate and examining the plates under U.V. light at 254nm. The areas corresponding to aldosterone were cut out and eluted with 5 ml of acetone. The eluates were evaporated and redissolved with 0.35 ml of borate buffer. For determination of recovery 0.1 ml of the eluates were transferred to counting vials and 0.4ml of borate buffer plus 10ml of the scintillation fluid were added to achieve counting conditions identical to those of the RIA samples. For determination of the aldosterone concentration preexisting in the antiserum 50~1 of a
80
1: 100 antiserum dilution were treated as described above but extraction with methylene chloride was repeated three times resulting in a complete extraction as checked by radioactive recovery. RZA procedure. Duplicate aliquots (0.1 ml) of the eluates and triplicate standards containing t&l000 pg of aldosterone in 0.1 ml of borate buffer were mixed with [3H]-aldosterone (10,00&15,000 d.p.m. = 16 24 pg in 0.5 ml of borate buffer) and antiserum diluted with y-globulin buffer (0.05 ml, final antiserum dilution in the 0.65 ml sample 1: 500,000). The tubes were mixed and incubated at 0” for 20 h. Separation of free and bound aldosterone was achieved by 0.05 ml of the Dextran-coated charcoal suspension simultaneously mixed with all assay tubes by shaking the coal from polyethylene caps fitted to the assay tubes. All samples were centrifuged simultaneously for 5 min at 0” and 2500 g. 0.5 ml-Aliquots of the supernatant together with 10ml of the scintillation fluid were transferred to counting vials and counted in a Mark II Liquid Scintillation Counter (Nuclear Chicago). Construction of the standard curve. “Absolute 3, bound’ were calculated according to the formula: B - NIT - N x 100. where B = c.p.m. bound, ‘Z = total c.p.m., N = non specific c.p.m. (all counts based on the 0.5 ml-aliquot out of 0.7 ml of supernatant) and plotted vs. the total pg of the standards on a logit/log paper. Total pg of the standards were calculated as the sum of pg of standard aldosterone added plus pg of [3H]-aldosterone (calculated from the S.A.) plus pg of aldosterone preexisting in the antiserum (5 pg/ml in the original antiserum corresponding to 6.5 pg/O.65 ml in the 1: 500,000 antiserum dilution). By this procedure a linear logit/log calibration curve was obtained in the range of 3S53Opg of total aldosterone, i.e., 0-5OOpg of standard aldosterone added (Fig. 1). Calculation of the unknowns. Since the total counts
-I
z
70
2 E : 5 p
60 50 10 30 20
pg oldo added
Fig. 1. Standard curve of the aldosterone radioimmunoassay calculated from the standards of 5 different assays performed during 1 week (rt f 2 SD., N = 15). “Absolute percent bound” are plotted on a lonit scale (ordinate) vs. the total amount of aldosterone (standard aldosterone added nlus DP of [%I-aldostkrone p/is pg of endogenous aldosterone preexisting in the antiserum) on a log-&ale (abscissa). Linear regression line (least square method): Y = -2.13 x + 2.53.
Radioimmunoassay for aldosterone in rats of the unknown samples were increased by the counts of the internal standard added for determination of individual recovery, it was necessary to compute “individual % bound”: (B - N) x 100: (T f 5/7 R - N), where B = c.p.m. of an unknown, N = non specific c.p.m., T = total c.p.m. added to the RIA-sample, R = c.p.m. recovered in the 0.1 ml-aiiquot of the respective unknown (R was reduced by 5/7, because all other c.p.m. were based on the 0.5 ml-aliquot out of 0.7 ml supernatant). Total pg of aldosterone were read of the standard curve, the pg of preexisting aidosterone (pg labeled aldosterone corresponding to the individual total counts plus pg endogenous aldosterone of the antiserum) were substracted and the netto pg obtained were corrected by the recovery of the respective 0.1 ml-aliquot of the unknown. A programable mini-computer (Hewlett-Packard, HP 65) was used for all calculations including logit transformation, linear regression and interpolation of the assay data. Plasma aldosterone concentrations were calculated as pg/mi and urinary aldosterone excretion rates were computed by multiplying urinary aldosterone concentration by the respective urine vol. Determination of the specific activity of [3H]-aldosterone by RIA. Increasing amounts of extra
[3H]-aldosterone were added to zero-standards and processed together with a usual standard curve (see Table 1). Individual total counts (samples with extra [3H]-aldosterone, but without charcoal separation) and individual non specific counts (samples with extra [3H]-aldosterone, but without antiserum) were also measured. Individual % bound of the samples containing extra [3H]-~dosterone were computed and the netto pg of aldosterone ~rresponding to extra [3H]-aldosterone were calculated (total pg read of the standard curve minus total pg of zero-standards). S.A. was calculated as pg of aldosterone per 1000d.p.m. at a counting efficiency of 49%. Mean S.A. of [1.2.6.7-3H]-aldosterone as measured by RIA was 1.60 pg/lOOOd.p.m. (= 98 Ci/mmol) as compared to 1.59 pg/lOOOd.p.m. (= 103 Ci/mmol) as specified by the manufacturer. RESULTS
Rudioactive recovery. Overall recovery (after extraction and t.1.c.)of the internal radioactive standard was 73.2 f 9.1% (X + S.D., N = 108; range 54-83x) from
849
rat plasma samples and was 76.6 + 4.0% (X k SD. N = 100; range 67-85x) from rat urine samples. Blank and sensitivity. Water blanks assayed in 10 different experiments gave a mean value of 10 + 7.5 pg/sample (X + S.D.) with a reading on the standard curve of O-4 pg/tube. The sensitivity was calculated from 5 standard curves (see Fig. 1) according to [S]; it was 5 pg at the 95% confidence level (2.145 S.D., N = 15). Therefore, the least-detectable dose was 25 pg/original sample. Accuracy. Aqueous standards of unlabeled aldosterone (5oOpg or 1OOOpg)were processed through the entire procedure in 13 different experiments. Recovery of unlabeled aldosterone @ 4 S.D.) was 104 + 6.5% (N = 6) and 97 + 6.9% (N = 7) respectively. Within-assay variation. Ten 3 ml samples of a pool’ of rat plasma (mean aldosterone concentration 63 pgjml) and then 1 ml samples of another pool of rat plasma (mean aidosterone concentration 197 pg/ml) were run in a single experiment. The coefficients of variation were 9.5% and 5.6x, respectively. In eight 1 ml samples of a pool of rat urine (mean aldosterone concentration 403 pg/ml) the coefficient of variation was 4.3%. Between-assay variation. 1 ml samples of a pool of rat plasma (mean aldosterone concentration 228 pg/ml) were assayed in 10 different experiments during a period of 3 months. The coefficient of variation was 8.876. When twenty-two 1 ml samples of urine of different rats (aldosterone concentrations ranging from 20%7OOpgjml) were determined in 2 different assays, the coefficient of variation was 6.9%. Spec~c~ty. Specificity is obtained by the chromatographic purifi~tion involved in the assay. As can be seen from Table 2 there is a definite separation of aldosterone from corticosterone, progesterone and 18-OH-deoxycorticosterone, which are the steroids most probable to crossreact with the antibody used [lo]. In plasma samples (4 ml) of rats adrenalectomized 48 h prior to sacrifice and in urine samples (6-8 ml) of rats adrenalectomized 7 days prior to sampling no aldosterone could be detected. In some experiments samples of plasma and urine obtained from rats with experimentally induced low and high aldosterone levels were assayed after an additional purification by a second t.1.c. using solvent system B. The values measured after the second t.1.c. were compared with those obtained after a single tic. using solvent
Table 1. Determination of the specific activity of [1.2.6.7-3H]-aldosterone by radioimmunoassay
Sample NO. 1 2 3 4
EXhXi [‘HI-aldo (c.p.m.) 25.362 38.715 51.147 66,360
Individual* total EoUntS (c.p.m.) 32,736 46,089 58.521 73.734
individual non specific CO”IlfS (c.p.m.) 210 339 454 557
GYlllts bound (c.p.m.) 4919 5185 5494 SS98
Individual % bound 14.5 10.6 8.7 7.3
Mass of [3H1-aldosterone total pg Netto 114 158 xnl 235
pgt
84 128 170 205
%A. PB atdo/ loo0 d.p.m.t 1.62 1.62 1.63 1.5t
* Individual total counts = c.p.m. of extra [3H]-afdosterone plus 7374c.p.m. of [3H]-aldosterone added to the RIA as tracer. tNetto pg corresponding to extra [3H]-aldosterone = total pg read of the standard curve minus total pg of zero standard (30 pg). $Calculation based on a counting efficiency of 49%.
850
MATHIA~HILFENHAUS Table 2. Separation of several steroids in the chromatographic
Steroid
Progesterone 11.Deoxycorticosterone Corticosterone Cortisone 18-OH-Deoxycorticosterone Cortisol Aldosterone 18.OH-Corticosterone
Solventsystem A (benzeneacetone_H~O. 3:2:0.025, v/v) R,-VtlbX
systems used in the aldosterone assay Solvent system B (ethylacetateacetone_H~O, 6: 1:0.025. v/v) R,-V&X
0.96 0.84 0.58 0.56 0.52
0.96 0.91 0.79 0.84 -
0.48 0.38 0.28
0.80 0.50 _
system A (Table 3). The differences were within the variability of the assay and no systematic deviation was observed (t > 1, paired t-test). Acid hydrolysis of rat urine. Four 1 ml-aliquots of a pool of rat urine were acidified to pH 1 with concentrated HCl and allowed to stand for 20 h at room temperature. Aldosterone concentration was determined and compared with unacidified samples of the same urine pool. Acid hydrolysis increased the concentration of urinary free aldosterone by 11% only. Stability of aldosterone in rat urine. Since urine of rats was sampled without any preservative at 24” over a period of 24 h, the stability of urinary aldosterone was checked. A loss of more than 10% was not observed until 4 days of exposure to these conditions. Normal values. Mean plasma aldosterone concentration (jz &-S.D., N = 6) in rats fed the standard diet Table 3. Comparison of aldosterone values obtained by the use of either a single t.1.c. (solvent system A) or of 2 subsequent t.1.c. (solvent system A followed by solvent system B)*
and sacrificed at 08.00 or at 18.00 was 40 f 17 pg/ml (range 23-72) and 153 + 76pg/ml (range 5%288), respectively. 24 h-Urinary aldosterone excretion (rt f S.D., N = 6) in these rats measured on two subsequent 5.0 f 0.9 ng/24 h days was and 4.5 f 0.7 ng/24 h (range 3.9-5.8), respectively. In rats under dietary sodium restriction for 2 weeks plasma aldosterone concentration (X k S.D., N = 3) was 9730 St 71Opg/ml at 08.00 and was 17,210 +_ 1440 pg/ml at 18.00. 24 h-urinary aldosterone excretion (X + S.D., N = 5) was 223 f 34 ng/24 h in these rats. DISCUSSION Since 1970 several RIA have been developed for the determination of aldosterone in biological fluids in man (for review see [ 1 l]), some of which have been also applied to rat plasma [l, 2,3,6], but not to rat urine. The RIA described in this study has been used
for the measurement of physiological amounts of aldosterone in plasma as well as in urine of rats and Plasma aldosterone has proven to be valid for both purposes. concentration Despite the rather high specificity of the antiserum PBW Treatment 1. t.1.c. 2. t.1.c. used [lo] the minute amount of aldosterone in biological fluids necessitates the separation of aldosterControl (08.00) 57 52 47 Control (08.00) 51 one from other corticosteroids before the binding 116 Control (18.00) 106 assay. In agreement with others [12] we found t.1.c. Control (18.00) 118 110 ACTH 250 j@kgt 365 327 to suit this purpose quite well, because it is rapid 479 456 ACTH 500 @Fgt Furosemide 25 mg/kgl 451 681 and simple. However, inclusion of a chromatographic Furosemide 25 mg/kgt 388 368 step demands the correction for procedural losses by individual recovery. When using internal radioactive Urinary aldosterone excr&on standards for determination of individual recovery the ng/2h b problem arises, how to correct for the radioactive 1. t.1.c. 2. t.1.c. aldosterone preexisting in the unknown samples subDOC-Depot 2 mgjkg$ 0.86 0.58 mitted to RIA. This problem has been solved [13] 0.84 0.61 DOC-Depot 2 m&k& Control 5.49 4.49 by calculating the bound fraction of the unknown Control 4.69 4.64 sample as percentage of the individual total counts ACTH 2 x 25 NJ/kg/day11 15.11 16.46 ACTH 2 x 25 IU/Lrg/dayll 15.79 16.11 of the same sample. In addition, this calculation Furosemide 25 mg/kg/dayT 20.71 21.74 method enables the determination of the specific acFurosemide 25 mg,kg/dafil 19.36 20.54 tivity of the tracer by RIA (Table 1). * Data from experiments published elsewhere in full Linearization of the standard curve was not detail [S]. t ACTH (SynacthenQ, Ciba) was injected i.v. 1 h obtained by logit transformation when total percent prior to blood sampling. $ Furosemide (Lasix@, Hoechst) binding was plotted vs. the log of inert aldosterone 25 mg/kg was injected SC. 90 min prior to blood sampling. 5 DOC-Depot (Cortiron-Depot@, Schering) 2 mg/kg was added. Even when the log of total amount of antigen injected S.C. 7 days prior to uiine sampling. 11ACTH (inert plus labeled antigen) was used as the dose par(Acethropan’x’, Hoechst) 2 x 25 IV/kg/day was injected S.C. for 2 days prior to urine sampling. 7 Furosemide (Lasix@, ameter [14], a slight flattening persisted at the upper part of the standard curve. Therefore, the aldosterone Hoechst) 1 x 25 mg/kg/day was injected S.C. for 7 days prior to urine sampling. content of the antiserum itself was determined after
851
Radioimmuno~say for aldosterone in rats extraction by methylene chloride and was found to be remarkably high (5 pg/ml) as has also been shown in rabbit antisera against several corticosteroids [lS]. Repeated treatment of the antiserum with Dextrancoated charcoal did not si~ific~tly reduce this high aldosterone concentration. Therefore, the amount of endogenous aldosterone of the antiserum was added to the total amount of antigen (“total pg”). This procedure resulted in a linearization of the calibration curve allowing the use of a mini-computer for calculation of all assay data. Specificity of the aldosterone assay is difficult to establish. No aldosterone was detected in plasma or urine of rats after adrenalectomy. However, those substances most probable to cross-react with the antibody are expected to originate from the adrenal cortex and therefore to be eliminated by adrenalectomy as well. On the other hand, several corticosteroids known to cross-react with the antibody used [lo] are clearly separated by the t.1.c. involved in our method (Table 2). Nevertheless, there may be steroids or metabolites, which are not identified thus far and hence can not be tested directly. Specificity of the assay is demonstrated more likely by the results obtained with plasma and urine samples purified by two subsequent t.1.c. using different solvent systems (Table 3). There was no systematic decrease in the aldosterone concentrations after the second t.1.c. and the differences between the values obtained after one or two t.1.c. were not si~i~cant. In man, urinary aldosterone excretion rate is usually measured after acid hydrolysis of the aldosterone-1%oxo-glucuronide. In the rat, acid-hydrolysable aldosterone metabolites and, in particular, aldosterone-l&oxo-giucuronide are excreted in the urine in minimal amounts only [l&17]. In augment with these findings aldosterone concentration increased by only ll%, when rat urine was subjected to acid hydrolysis. Since the RIA used is sensitive to detect free aldosterone in small samples of urine, hydrolysis of rat urine was omitted and aldosterone excretion rates were based on the dete~ination of urinary free aldosterone. Normal values of plasma aldosterone concentration in male Sprague-Dawley rats showing a pronounced circadian rhythm [4,5,7] ranged from 23 to 288 pg/ml. when rats were kept on a standard diet, and from 7000 to 2O,~p~ml, when rats were kept on a sodium-poor diet for 14 days. These values are in good agreement with those reported in the literature for male adult rats regardless whether an isotope derivative method [IS] or a radioimmunological method [2,3,6] was used. On the other hand, our method shows that normal values of urinary aldosterone excretion are lower than levels of
100-380 rig/24 h [IS] or 40-50 ng/24 h [20] reported previously for rats. The differences obtained appear to result from the improved isolation technique and the use of RIA instead of polarography [19] or double isotope derivative technique [ZO] for measurement of aldosterone. The determination of urinary aldosterone excretion rate is a useful parameter when investigating the control of aldosterone secretion in intact, unstressed rats. The biological reliability of this parameter has been established in experiments in rats using DOC, ACTH, furosemide and changes in sodium balance to alter the activity of the renin-angiotensin-aldosterone system [8]. Acknowledgements-The author is indebted to Dr. P. Vecsei, Ph~makoIo~sch~ Institut der Universitlt Heideiberg, for the kind gift of the antiserum against aldosterone-21-hemisuccinate (sheep 3) and for helpful advice. The excellent technical assistance of Miss E. Hempen is gratefully acknowledged. REFERENCES 1. Campbell W. B., Brooks S. N. and Pettinger W. A.: Science 184 (1974) 994-996. 2. Campbell W. B., Pettinger W. A., Keeton K. and Brooks S. N.: J. Pharmac. exo. Ther. 193 11975) I , 166175. Kubo S., Ganten D., Ganten U., Nowaczynski W. and Genest J.: Endocrinology 94 (1974) 459-463. Hi~enhaus M. : ~uun~-Sch~edebera~ - Arch. Phur~aco~. Suppl. 282 (1974) R 32. Hilfenhaus M.: int. J. Chronobiol. 3 (1976) 213-229.
Dietz R., Mast G. J., Vecsei P., Gless K. H., Oster P. and Gross F.: Acta endocr.., Cooenh. 79 (1975) . 317-328. 7. Hilfenhaus M.: Archs Toxicol. 36 (1976) 305316. 8. Hilfenhaus M.: Acta endocr., Copenh. 85 (1977) 134-142. 9. Feldman H. and Rodbard D.: In ~~nc~Dfes of Comvetitive ~ote~n-3i~ing Assays (Edited by W.* D. ddell and W. Daughaday). J. B. Lippincott, Philadelphia (1971) p. 164. 10. Vecsei P., Gless K. H., Winter M. and HankaPostozky M.: Acta endocr., Copenh. 74 (1973) 307-315. 11. Vecsei P. and Gless K. H.: Aldosterone Radioimrn~~~~uy. Enke Verlag, Stuttgart (1975). 12. Spat A. and Jozan S.: J. steroid ~~ochem. 4 (1973) so9518 13. Cekan 2.: J. steroid Biochem. 6 (1975) 271-275. 14. Gilman A. G.: Proc. natn. Acad. Sci. U.S.A. 67 (1970) 305-312. 15. Gless K. H., Hanka M., Vecsei P. and Gross F.: Acta endocr., Copenh. 75 (1974) 342-349. 16. McCaa C. S. and Sulya L. L.: ~~ocr~~oloay __ 79 0966) . , 815-818. 17. Mohring J., Mohring B. and Siegenthaler W.: KEn. Wschr. 46 (1968) 22-24. 18. Bojesen E.: Eur: J. Steroids 1 (1966) 145-169. 19. Paroli E. and Melchiorri P.: Biochem. Pharmac. 6 (1961) 1-17. 20. Schwartz J., Bloch R. and Velly J.: Canud. med. Ass. J. 90 (1964) 24>244.
ofSterotd
Biochemisrry,
1977, Vol. 8. pp. 847-851. Pergamon Press. Printedin Great Britain
EVALUATION OF RADIOIMMUNOASSAY ALDOSTERONE IN URINE AND PLASMA OF RATS*
FOR
MATHIAS HILFENHAUS Institut
fur Pharmakologie
der Medizinischen Hochschule Hannover, Karl-Wiechert-Allee D 3000 Hannover 61, Germany
9,
SUMMARY
A method is described for the measurement of aldosterone in urine and plasma of rats and the sensitivity, specificity, accuracy and precision of the assay are assessed. The method includes one extraction step by methylene chloride, purification of the extract by one thin layer chromatography and determination of the purified aldosterone by radioimmunoassay. [3H]-Aldosterone is used for the radioimmunoassay as well as for the determination of the recovery of the individual sample. A modified calculation of the radioimmunoassay results is described to correct for the varying amounts of [3H]-aldosterone of the recovery preexisting in the unknowns but not in the standards of the calibration curve. Normal values of plasma aldosterone concentration in adult male Sprague-Dawley rats (23-288 pg/ml depending on the time of day) as measured with our assay are in good agreement with the values reported in the literature. Normal values of urinary aldosterone excretion rate in these rats (3.95.8 ng/24 h) are much lower than those reported in the literature possibly due to the higher specificity of the present assay.
INTRODUCTION
In the rat, in uiuo studies on aldosterone
have been restricted for a long time to the determination of aldosterone secretion in surgically stressed animals (for review see Cl]). Due to the development of highly sensitive radioimmunoassays (RIA) the determination of physiological concentrations of aldosterone in the blood of intact, i.e., conscious and unstressed rats has become possible [l-6]. We recently also reported on aldosterone excretion rates in the urine of rats [7,8]. Since data on urinary aldosterone in the rat as measured by RlA have not been published yet, we herein describe a RIA for the measurement of aldosterone in urine and plasma of rats using extraction and thin layer chromatography (t.1.c.) for separation of aldosterone. This assay has been used routinely in our laboratory during the last 4 years and has proven to be comparatively rapid and simple. MATERIALS
AND METHODS
Animals. Male Sprague-Dawley rats (20&250 g) were kept in a room with controlled temperature (23-24”) and humidity (>SO%) under a lighting regime of 12 h light (06.0&18.00) and 12 h dark (18.0&06.00). The animals were maintained on a standard diet (Na+ 96 mmol/kg, K+ 143 mmol/kg; Altromin. Lage, Germany) and tap water ad libitum or on * Supported by the Deutsche Forschungsgemeinschaft (Sonderforschunasbereich 146 “Versuchstierforschuna”. G 5). Requests f& reprints should be addressed to: Dr: M. Hilfenhaus, Institut fur Pharmakologie, Medizinische Hochschule Hannover, Karl-Wiechert-Allee 9, D 3000 Hannover 61, Germany.
a low-sodium diet (Na+ 4.4mmol/kg, K+ 60mmol/kg; Altromin) and demineralized water ad libitum. For determination of plasma aldosterone rats were decapitated without anaesthesia and blood was collected from the trunk using EDTA (1 mg/ml) as. anticoagulant. Plasma was separated by centrifugation at 0” and stored at -20” until assayed. For measurement of urinary aldosterone excretion rats were placed in individual metabolic cages made of Plexiglas with free access to food and water, the urine--separated from feces and spilled food-was collected in 25 ml-Erlenmeyer flasks. After that urine samples were stored at -20” until assayed. Plasma corticosterone concentrations were identical in rats kept in usual cages or kept in metabolic cages indicating that sitting in a metabolic cage was no stress for the animals. Materials. Methylene chloride, benzene, acetone (all technical grade) were obtained from Merck (Darmstadt) and distilled over a 120cm packed column filled with wire-mesh rings. Toluene (technical grade, Merck) and Triton-X-100 (technical grade, Serva, Heidelberg) were used without further purification. Precoated t.1.c. plates containing a fluorescence indicator (DC-Alufolien, Kieselgel F 254, Merck) were used as purchased. Chromatography was performed in the solvent system benzene-acetone-H,0 (3:2:0.025 by vol) (system A) or in the system ethylacetate-acetone-H,0 (6:1:0.025 by vol) (system B). [1.2.6.7-3H]-Aldosterone (S.A. 103 Ci/mmol, Radiochemical Centre, Amersham) was stored in benzene:ethanol (9: 1 by vol) at 4”. Radiochemical purity was checked by t.1.c. at 3 months intervals. A measureable decomposition of 5% was observed after 6 months of storage and the tracer was purified by t.1.c. 84-l
848
MATHW HILFENHAUS
using solvent system A. o-aldosterone was kindly supplied by Ciba and was used without further purification. The antiserum against aldosterone-21-hemisuccinate was raised in a sheep [lo]. Borate buffer (0.05 M, pH 8), y-globulin buffer (borate buffer containing 1% of bovine immunoglobulin, Behring, Marburg) and Dextran-coated charcoal (6g Norit A, Serva, and 0.2 g Dextran T 70, Pharmacia, Uppsala, suspended in lOOm1 of distilled water) were stored at 4” and were stable for at least 1 month. The scintillation fluid consisted of 40g PPO (2.5-diphenyloxazole, Merck), 1 g POPOP ‘(l&bis-2-(5phenyloxazolyl)-benzene, Merck), 2 1 of Triton-X-100 and 8 1 of toluene. Extraction and purijkation of plasma and urine samples. 0.05-4 ml of plasma or urine were mixed with
known amounts of [3H]-aldosterone (about 5000d.p.m. = 8 pg) as internal standard for determination of individual recovery. After an equilibration period of 12 h the samples were extracted with 20ml of methylene chloride by magnetic stirring for 15 min. The aqueous layer was discarded and the filtered organic extract evaporated. The dry extracts were dissolved in acetone (3 x 0.2 ml) and transferred to the t.1.c. plates using a Desaga-autoliner. The plates were developed in the solvent system A. Aldosterone was located by running spots of unlabeled aldosterone (about 0.5 pg) with each plate and examining the plates under U.V. light at 254nm. The areas corresponding to aldosterone were cut out and eluted with 5 ml of acetone. The eluates were evaporated and redissolved with 0.35 ml of borate buffer. For determination of recovery 0.1 ml of the eluates were transferred to counting vials and 0.4ml of borate buffer plus 10ml of the scintillation fluid were added to achieve counting conditions identical to those of the RIA samples. For determination of the aldosterone concentration preexisting in the antiserum 50~1 of a
80
1: 100 antiserum dilution were treated as described above but extraction with methylene chloride was repeated three times resulting in a complete extraction as checked by radioactive recovery. RZA procedure. Duplicate aliquots (0.1 ml) of the eluates and triplicate standards containing t&l000 pg of aldosterone in 0.1 ml of borate buffer were mixed with [3H]-aldosterone (10,00&15,000 d.p.m. = 16 24 pg in 0.5 ml of borate buffer) and antiserum diluted with y-globulin buffer (0.05 ml, final antiserum dilution in the 0.65 ml sample 1: 500,000). The tubes were mixed and incubated at 0” for 20 h. Separation of free and bound aldosterone was achieved by 0.05 ml of the Dextran-coated charcoal suspension simultaneously mixed with all assay tubes by shaking the coal from polyethylene caps fitted to the assay tubes. All samples were centrifuged simultaneously for 5 min at 0” and 2500 g. 0.5 ml-Aliquots of the supernatant together with 10ml of the scintillation fluid were transferred to counting vials and counted in a Mark II Liquid Scintillation Counter (Nuclear Chicago). Construction of the standard curve. “Absolute 3, bound’ were calculated according to the formula: B - NIT - N x 100. where B = c.p.m. bound, ‘Z = total c.p.m., N = non specific c.p.m. (all counts based on the 0.5 ml-aliquot out of 0.7 ml of supernatant) and plotted vs. the total pg of the standards on a logit/log paper. Total pg of the standards were calculated as the sum of pg of standard aldosterone added plus pg of [3H]-aldosterone (calculated from the S.A.) plus pg of aldosterone preexisting in the antiserum (5 pg/ml in the original antiserum corresponding to 6.5 pg/O.65 ml in the 1: 500,000 antiserum dilution). By this procedure a linear logit/log calibration curve was obtained in the range of 3S53Opg of total aldosterone, i.e., 0-5OOpg of standard aldosterone added (Fig. 1). Calculation of the unknowns. Since the total counts
-I
z
70
2 E : 5 p
60 50 10 30 20
pg oldo added
Fig. 1. Standard curve of the aldosterone radioimmunoassay calculated from the standards of 5 different assays performed during 1 week (rt f 2 SD., N = 15). “Absolute percent bound” are plotted on a lonit scale (ordinate) vs. the total amount of aldosterone (standard aldosterone added nlus DP of [%I-aldostkrone p/is pg of endogenous aldosterone preexisting in the antiserum) on a log-&ale (abscissa). Linear regression line (least square method): Y = -2.13 x + 2.53.
Radioimmunoassay for aldosterone in rats of the unknown samples were increased by the counts of the internal standard added for determination of individual recovery, it was necessary to compute “individual % bound”: (B - N) x 100: (T f 5/7 R - N), where B = c.p.m. of an unknown, N = non specific c.p.m., T = total c.p.m. added to the RIA-sample, R = c.p.m. recovered in the 0.1 ml-aiiquot of the respective unknown (R was reduced by 5/7, because all other c.p.m. were based on the 0.5 ml-aliquot out of 0.7 ml supernatant). Total pg of aldosterone were read of the standard curve, the pg of preexisting aidosterone (pg labeled aldosterone corresponding to the individual total counts plus pg endogenous aldosterone of the antiserum) were substracted and the netto pg obtained were corrected by the recovery of the respective 0.1 ml-aliquot of the unknown. A programable mini-computer (Hewlett-Packard, HP 65) was used for all calculations including logit transformation, linear regression and interpolation of the assay data. Plasma aldosterone concentrations were calculated as pg/mi and urinary aldosterone excretion rates were computed by multiplying urinary aldosterone concentration by the respective urine vol. Determination of the specific activity of [3H]-aldosterone by RIA. Increasing amounts of extra
[3H]-aldosterone were added to zero-standards and processed together with a usual standard curve (see Table 1). Individual total counts (samples with extra [3H]-aldosterone, but without charcoal separation) and individual non specific counts (samples with extra [3H]-aldosterone, but without antiserum) were also measured. Individual % bound of the samples containing extra [3H]-~dosterone were computed and the netto pg of aldosterone ~rresponding to extra [3H]-aldosterone were calculated (total pg read of the standard curve minus total pg of zero-standards). S.A. was calculated as pg of aldosterone per 1000d.p.m. at a counting efficiency of 49%. Mean S.A. of [1.2.6.7-3H]-aldosterone as measured by RIA was 1.60 pg/lOOOd.p.m. (= 98 Ci/mmol) as compared to 1.59 pg/lOOOd.p.m. (= 103 Ci/mmol) as specified by the manufacturer. RESULTS
Rudioactive recovery. Overall recovery (after extraction and t.1.c.)of the internal radioactive standard was 73.2 f 9.1% (X + S.D., N = 108; range 54-83x) from
849
rat plasma samples and was 76.6 + 4.0% (X k SD. N = 100; range 67-85x) from rat urine samples. Blank and sensitivity. Water blanks assayed in 10 different experiments gave a mean value of 10 + 7.5 pg/sample (X + S.D.) with a reading on the standard curve of O-4 pg/tube. The sensitivity was calculated from 5 standard curves (see Fig. 1) according to [S]; it was 5 pg at the 95% confidence level (2.145 S.D., N = 15). Therefore, the least-detectable dose was 25 pg/original sample. Accuracy. Aqueous standards of unlabeled aldosterone (5oOpg or 1OOOpg)were processed through the entire procedure in 13 different experiments. Recovery of unlabeled aldosterone @ 4 S.D.) was 104 + 6.5% (N = 6) and 97 + 6.9% (N = 7) respectively. Within-assay variation. Ten 3 ml samples of a pool’ of rat plasma (mean aldosterone concentration 63 pgjml) and then 1 ml samples of another pool of rat plasma (mean aidosterone concentration 197 pg/ml) were run in a single experiment. The coefficients of variation were 9.5% and 5.6x, respectively. In eight 1 ml samples of a pool of rat urine (mean aldosterone concentration 403 pg/ml) the coefficient of variation was 4.3%. Between-assay variation. 1 ml samples of a pool of rat plasma (mean aldosterone concentration 228 pg/ml) were assayed in 10 different experiments during a period of 3 months. The coefficient of variation was 8.876. When twenty-two 1 ml samples of urine of different rats (aldosterone concentrations ranging from 20%7OOpgjml) were determined in 2 different assays, the coefficient of variation was 6.9%. Spec~c~ty. Specificity is obtained by the chromatographic purifi~tion involved in the assay. As can be seen from Table 2 there is a definite separation of aldosterone from corticosterone, progesterone and 18-OH-deoxycorticosterone, which are the steroids most probable to crossreact with the antibody used [lo]. In plasma samples (4 ml) of rats adrenalectomized 48 h prior to sacrifice and in urine samples (6-8 ml) of rats adrenalectomized 7 days prior to sampling no aldosterone could be detected. In some experiments samples of plasma and urine obtained from rats with experimentally induced low and high aldosterone levels were assayed after an additional purification by a second t.1.c. using solvent system B. The values measured after the second t.1.c. were compared with those obtained after a single tic. using solvent
Table 1. Determination of the specific activity of [1.2.6.7-3H]-aldosterone by radioimmunoassay
Sample NO. 1 2 3 4
EXhXi [‘HI-aldo (c.p.m.) 25.362 38.715 51.147 66,360
Individual* total EoUntS (c.p.m.) 32,736 46,089 58.521 73.734
individual non specific CO”IlfS (c.p.m.) 210 339 454 557
GYlllts bound (c.p.m.) 4919 5185 5494 SS98
Individual % bound 14.5 10.6 8.7 7.3
Mass of [3H1-aldosterone total pg Netto 114 158 xnl 235
pgt
84 128 170 205
%A. PB atdo/ loo0 d.p.m.t 1.62 1.62 1.63 1.5t
* Individual total counts = c.p.m. of extra [3H]-afdosterone plus 7374c.p.m. of [3H]-aldosterone added to the RIA as tracer. tNetto pg corresponding to extra [3H]-aldosterone = total pg read of the standard curve minus total pg of zero standard (30 pg). $Calculation based on a counting efficiency of 49%.
850
MATHIA~HILFENHAUS Table 2. Separation of several steroids in the chromatographic
Steroid
Progesterone 11.Deoxycorticosterone Corticosterone Cortisone 18-OH-Deoxycorticosterone Cortisol Aldosterone 18.OH-Corticosterone
Solventsystem A (benzeneacetone_H~O. 3:2:0.025, v/v) R,-VtlbX
systems used in the aldosterone assay Solvent system B (ethylacetateacetone_H~O, 6: 1:0.025. v/v) R,-V&X
0.96 0.84 0.58 0.56 0.52
0.96 0.91 0.79 0.84 -
0.48 0.38 0.28
0.80 0.50 _
system A (Table 3). The differences were within the variability of the assay and no systematic deviation was observed (t > 1, paired t-test). Acid hydrolysis of rat urine. Four 1 ml-aliquots of a pool of rat urine were acidified to pH 1 with concentrated HCl and allowed to stand for 20 h at room temperature. Aldosterone concentration was determined and compared with unacidified samples of the same urine pool. Acid hydrolysis increased the concentration of urinary free aldosterone by 11% only. Stability of aldosterone in rat urine. Since urine of rats was sampled without any preservative at 24” over a period of 24 h, the stability of urinary aldosterone was checked. A loss of more than 10% was not observed until 4 days of exposure to these conditions. Normal values. Mean plasma aldosterone concentration (jz &-S.D., N = 6) in rats fed the standard diet Table 3. Comparison of aldosterone values obtained by the use of either a single t.1.c. (solvent system A) or of 2 subsequent t.1.c. (solvent system A followed by solvent system B)*
and sacrificed at 08.00 or at 18.00 was 40 f 17 pg/ml (range 23-72) and 153 + 76pg/ml (range 5%288), respectively. 24 h-Urinary aldosterone excretion (rt f S.D., N = 6) in these rats measured on two subsequent 5.0 f 0.9 ng/24 h days was and 4.5 f 0.7 ng/24 h (range 3.9-5.8), respectively. In rats under dietary sodium restriction for 2 weeks plasma aldosterone concentration (X k S.D., N = 3) was 9730 St 71Opg/ml at 08.00 and was 17,210 +_ 1440 pg/ml at 18.00. 24 h-urinary aldosterone excretion (X + S.D., N = 5) was 223 f 34 ng/24 h in these rats. DISCUSSION Since 1970 several RIA have been developed for the determination of aldosterone in biological fluids in man (for review see [ 1 l]), some of which have been also applied to rat plasma [l, 2,3,6], but not to rat urine. The RIA described in this study has been used
for the measurement of physiological amounts of aldosterone in plasma as well as in urine of rats and Plasma aldosterone has proven to be valid for both purposes. concentration Despite the rather high specificity of the antiserum PBW Treatment 1. t.1.c. 2. t.1.c. used [lo] the minute amount of aldosterone in biological fluids necessitates the separation of aldosterControl (08.00) 57 52 47 Control (08.00) 51 one from other corticosteroids before the binding 116 Control (18.00) 106 assay. In agreement with others [12] we found t.1.c. Control (18.00) 118 110 ACTH 250 j@kgt 365 327 to suit this purpose quite well, because it is rapid 479 456 ACTH 500 @Fgt Furosemide 25 mg/kgl 451 681 and simple. However, inclusion of a chromatographic Furosemide 25 mg/kgt 388 368 step demands the correction for procedural losses by individual recovery. When using internal radioactive Urinary aldosterone excr&on standards for determination of individual recovery the ng/2h b problem arises, how to correct for the radioactive 1. t.1.c. 2. t.1.c. aldosterone preexisting in the unknown samples subDOC-Depot 2 mgjkg$ 0.86 0.58 mitted to RIA. This problem has been solved [13] 0.84 0.61 DOC-Depot 2 m&k& Control 5.49 4.49 by calculating the bound fraction of the unknown Control 4.69 4.64 sample as percentage of the individual total counts ACTH 2 x 25 NJ/kg/day11 15.11 16.46 ACTH 2 x 25 IU/Lrg/dayll 15.79 16.11 of the same sample. In addition, this calculation Furosemide 25 mg/kg/dayT 20.71 21.74 method enables the determination of the specific acFurosemide 25 mg,kg/dafil 19.36 20.54 tivity of the tracer by RIA (Table 1). * Data from experiments published elsewhere in full Linearization of the standard curve was not detail [S]. t ACTH (SynacthenQ, Ciba) was injected i.v. 1 h obtained by logit transformation when total percent prior to blood sampling. $ Furosemide (Lasix@, Hoechst) binding was plotted vs. the log of inert aldosterone 25 mg/kg was injected SC. 90 min prior to blood sampling. 5 DOC-Depot (Cortiron-Depot@, Schering) 2 mg/kg was added. Even when the log of total amount of antigen injected S.C. 7 days prior to uiine sampling. 11ACTH (inert plus labeled antigen) was used as the dose par(Acethropan’x’, Hoechst) 2 x 25 IV/kg/day was injected S.C. for 2 days prior to urine sampling. 7 Furosemide (Lasix@, ameter [14], a slight flattening persisted at the upper part of the standard curve. Therefore, the aldosterone Hoechst) 1 x 25 mg/kg/day was injected S.C. for 7 days prior to urine sampling. content of the antiserum itself was determined after
851
Radioimmuno~say for aldosterone in rats extraction by methylene chloride and was found to be remarkably high (5 pg/ml) as has also been shown in rabbit antisera against several corticosteroids [lS]. Repeated treatment of the antiserum with Dextrancoated charcoal did not si~ific~tly reduce this high aldosterone concentration. Therefore, the amount of endogenous aldosterone of the antiserum was added to the total amount of antigen (“total pg”). This procedure resulted in a linearization of the calibration curve allowing the use of a mini-computer for calculation of all assay data. Specificity of the aldosterone assay is difficult to establish. No aldosterone was detected in plasma or urine of rats after adrenalectomy. However, those substances most probable to cross-react with the antibody are expected to originate from the adrenal cortex and therefore to be eliminated by adrenalectomy as well. On the other hand, several corticosteroids known to cross-react with the antibody used [lo] are clearly separated by the t.1.c. involved in our method (Table 2). Nevertheless, there may be steroids or metabolites, which are not identified thus far and hence can not be tested directly. Specificity of the assay is demonstrated more likely by the results obtained with plasma and urine samples purified by two subsequent t.1.c. using different solvent systems (Table 3). There was no systematic decrease in the aldosterone concentrations after the second t.1.c. and the differences between the values obtained after one or two t.1.c. were not si~i~cant. In man, urinary aldosterone excretion rate is usually measured after acid hydrolysis of the aldosterone-1%oxo-glucuronide. In the rat, acid-hydrolysable aldosterone metabolites and, in particular, aldosterone-l&oxo-giucuronide are excreted in the urine in minimal amounts only [l&17]. In augment with these findings aldosterone concentration increased by only ll%, when rat urine was subjected to acid hydrolysis. Since the RIA used is sensitive to detect free aldosterone in small samples of urine, hydrolysis of rat urine was omitted and aldosterone excretion rates were based on the dete~ination of urinary free aldosterone. Normal values of plasma aldosterone concentration in male Sprague-Dawley rats showing a pronounced circadian rhythm [4,5,7] ranged from 23 to 288 pg/ml. when rats were kept on a standard diet, and from 7000 to 2O,~p~ml, when rats were kept on a sodium-poor diet for 14 days. These values are in good agreement with those reported in the literature for male adult rats regardless whether an isotope derivative method [IS] or a radioimmunological method [2,3,6] was used. On the other hand, our method shows that normal values of urinary aldosterone excretion are lower than levels of
100-380 rig/24 h [IS] or 40-50 ng/24 h [20] reported previously for rats. The differences obtained appear to result from the improved isolation technique and the use of RIA instead of polarography [19] or double isotope derivative technique [ZO] for measurement of aldosterone. The determination of urinary aldosterone excretion rate is a useful parameter when investigating the control of aldosterone secretion in intact, unstressed rats. The biological reliability of this parameter has been established in experiments in rats using DOC, ACTH, furosemide and changes in sodium balance to alter the activity of the renin-angiotensin-aldosterone system [8]. Acknowledgements-The author is indebted to Dr. P. Vecsei, Ph~makoIo~sch~ Institut der Universitlt Heideiberg, for the kind gift of the antiserum against aldosterone-21-hemisuccinate (sheep 3) and for helpful advice. The excellent technical assistance of Miss E. Hempen is gratefully acknowledged. REFERENCES 1. Campbell W. B., Brooks S. N. and Pettinger W. A.: Science 184 (1974) 994-996. 2. Campbell W. B., Pettinger W. A., Keeton K. and Brooks S. N.: J. Pharmac. exo. Ther. 193 11975) I , 166175. Kubo S., Ganten D., Ganten U., Nowaczynski W. and Genest J.: Endocrinology 94 (1974) 459-463. Hi~enhaus M. : ~uun~-Sch~edebera~ - Arch. Phur~aco~. Suppl. 282 (1974) R 32. Hilfenhaus M.: int. J. Chronobiol. 3 (1976) 213-229.
Dietz R., Mast G. J., Vecsei P., Gless K. H., Oster P. and Gross F.: Acta endocr.., Cooenh. 79 (1975) . 317-328. 7. Hilfenhaus M.: Archs Toxicol. 36 (1976) 305316. 8. Hilfenhaus M.: Acta endocr., Copenh. 85 (1977) 134-142. 9. Feldman H. and Rodbard D.: In ~~nc~Dfes of Comvetitive ~ote~n-3i~ing Assays (Edited by W.* D. ddell and W. Daughaday). J. B. Lippincott, Philadelphia (1971) p. 164. 10. Vecsei P., Gless K. H., Winter M. and HankaPostozky M.: Acta endocr., Copenh. 74 (1973) 307-315. 11. Vecsei P. and Gless K. H.: Aldosterone Radioimrn~~~~uy. Enke Verlag, Stuttgart (1975). 12. Spat A. and Jozan S.: J. steroid ~~ochem. 4 (1973) so9518 13. Cekan 2.: J. steroid Biochem. 6 (1975) 271-275. 14. Gilman A. G.: Proc. natn. Acad. Sci. U.S.A. 67 (1970) 305-312. 15. Gless K. H., Hanka M., Vecsei P. and Gross F.: Acta endocr., Copenh. 75 (1974) 342-349. 16. McCaa C. S. and Sulya L. L.: ~~ocr~~oloay __ 79 0966) . , 815-818. 17. Mohring J., Mohring B. and Siegenthaler W.: KEn. Wschr. 46 (1968) 22-24. 18. Bojesen E.: Eur: J. Steroids 1 (1966) 145-169. 19. Paroli E. and Melchiorri P.: Biochem. Pharmac. 6 (1961) 1-17. 20. Schwartz J., Bloch R. and Velly J.: Canud. med. Ass. J. 90 (1964) 24>244.
